// Author: Miguel de Icaza (miguel@gnu.org)
// Marek Safar (marek.safar@seznam.cz)
//
-// Licensed under the terms of the GNU GPL
+// Dual licensed under the terms of the MIT X11 or GNU GPL
//
-// (C) 2001, 2002 Ximian, Inc (http://www.ximian.com)
-// (C) 2004 Novell, Inc
+// Copyright 2001, 2002 Ximian, Inc (http://www.ximian.com)
+// Copyright 2004-2008 Novell, Inc
//
//
{
SeekableStreamReader reader;
SourceFile ref_name;
- SourceFile file_name;
+ CompilationUnit file_name;
+ bool hidden = false;
int ref_line = 1;
int line = 1;
int col = 0;
int current_token;
bool handle_get_set = false;
bool handle_remove_add = false;
- bool handle_assembly = false;
- bool handle_constraints = false;
+ bool handle_where = false;
bool handle_typeof = false;
- bool linq;
+ bool lambda_arguments_parsing;
Location current_location;
Location current_comment_location = Location.Null;
- ArrayList escapedIdentifiers = new ArrayList ();
+ ArrayList escaped_identifiers = new ArrayList ();
+
+ //
+ // Used mainly for parser optimizations. Some expressions for instance
+ // can appear only in block (including initializer, base initializer)
+ // scope only
+ //
+ public int parsing_block;
+ internal int query_parsing;
+
+ //
+ // Set when parsing generic declaration (type or method header)
+ //
+ public bool parsing_generic_declaration;
+
+ //
+ // The value indicates that we have not reach any declaration or
+ // namespace yet
+ //
+ public int parsing_declaration;
+ //
+ // The special character to inject on streams to trigger the EXPRESSION_PARSE
+ // token to be returned. It just happens to be a Unicode character that
+ // would never be part of a program (can not be an identifier).
+ //
+ // This character is only tested just before the tokenizer is about to report
+ // an error; So on the regular operation mode, this addition will have no
+ // impact on the tokenizer's performance.
+ //
+
+ public const int EvalStatementParserCharacter = 0x2190; // Unicode Left Arrow
+ public const int EvalCompilationUnitParserCharacter = 0x2191; // Unicode Arrow
+ public const int EvalUsingDeclarationsParserCharacter = 0x2192; // Unicode Arrow
+
//
// XML documentation buffer. The save point is used to divide
// comments on types and comments on members.
//
// See comment on XmlCommentState enumeration.
//
- XmlCommentState xmlDocState = XmlCommentState.Allowed;
+ XmlCommentState xml_doc_state = XmlCommentState.Allowed;
//
// Whether tokens have been seen on this line
//
bool any_token_seen = false;
- static Hashtable tokenValues;
+ static Hashtable token_values;
static readonly char[] simple_whitespaces = new char[] { ' ', '\t' };
private static Hashtable TokenValueName
{
get {
- if (tokenValues == null)
- tokenValues = GetTokenValueNameHash ();
+ if (token_values == null)
+ token_values = GetTokenValueNameHash ();
- return tokenValues;
+ return token_values;
}
}
}
public bool PropertyParsing {
- get {
- return handle_get_set;
- }
-
- set {
- handle_get_set = value;
- }
+ get { return handle_get_set; }
+ set { handle_get_set = value; }
}
- public bool AssemblyTargetParsing {
- get {
- return handle_assembly;
- }
-
- set {
- handle_assembly = value;
- }
- }
-
public bool EventParsing {
- get {
- return handle_remove_add;
- }
-
- set {
- handle_remove_add = value;
- }
+ get { return handle_remove_add; }
+ set { handle_remove_add = value; }
}
public bool ConstraintsParsing {
- get {
- return handle_constraints;
- }
-
- set {
- handle_constraints = value;
- }
+ get { return handle_where; }
+ set { handle_where = value; }
}
public bool TypeOfParsing {
- get {
- return handle_typeof;
- }
-
- set {
- handle_typeof = value;
- }
+ get { return handle_typeof; }
+ set { handle_typeof = value; }
}
-
+
public XmlCommentState doc_state {
- get { return xmlDocState; }
+ get { return xml_doc_state; }
set {
if (value == XmlCommentState.Allowed) {
check_incorrect_doc_comment ();
reset_doc_comment ();
}
- xmlDocState = value;
+ xml_doc_state = value;
}
}
public bool IsEscapedIdentifier (Location loc)
{
- foreach (LocatedToken lt in escapedIdentifiers)
+ foreach (LocatedToken lt in escaped_identifiers)
if (lt.Location.Equals (loc))
return true;
return false;
// Class variables
//
static CharArrayHashtable[] keywords;
- static Hashtable keywordStrings = new Hashtable ();
+ static Hashtable keyword_strings;
static NumberStyles styles;
static NumberFormatInfo csharp_format_info;
//
// Values for the associated token returned
//
- int putback_char;
+ internal int putback_char; // Used by repl only
Object val;
//
// Pre-processor
//
- Hashtable defines;
-
const int TAKING = 1;
const int ELSE_SEEN = 4;
const int PARENT_TAKING = 8;
Stack position_stack = new Stack (2);
class Position {
public int position;
+ public int line;
public int ref_line;
public int col;
+ public bool hidden;
public int putback_char;
public int previous_col;
public Stack ifstack;
public Position (Tokenizer t)
{
position = t.reader.Position;
+ line = t.line;
ref_line = t.ref_line;
col = t.col;
+ hidden = t.hidden;
putback_char = t.putback_char;
previous_col = t.previous_col;
if (t.ifstack != null && t.ifstack.Count != 0)
reader.Position = p.position;
ref_line = p.ref_line;
+ line = p.line;
col = p.col;
+ hidden = p.hidden;
putback_char = p.putback_char;
previous_col = p.previous_col;
ifstack = p.ifstack;
position_stack.Pop ();
}
- static void AddKeyword (string kw, int token) {
- keywordStrings.Add (kw, kw);
+ static void AddKeyword (string kw, int token)
+ {
+ keyword_strings.Add (kw, kw);
if (keywords [kw.Length] == null) {
keywords [kw.Length] = new CharArrayHashtable (kw.Length);
}
static void InitTokens ()
{
+ keyword_strings = new Hashtable ();
keywords = new CharArrayHashtable [64];
AddKeyword ("__arglist", Token.ARGLIST);
AddKeyword ("abstract", Token.ABSTRACT);
AddKeyword ("as", Token.AS);
AddKeyword ("add", Token.ADD);
- AddKeyword ("assembly", Token.ASSEMBLY);
AddKeyword ("base", Token.BASE);
AddKeyword ("bool", Token.BOOL);
AddKeyword ("break", Token.BREAK);
AddKeyword ("volatile", Token.VOLATILE);
AddKeyword ("while", Token.WHILE);
AddKeyword ("partial", Token.PARTIAL);
-#if GMCS_SOURCE
AddKeyword ("where", Token.WHERE);
-#endif
+
+ // LINQ keywords
+ AddKeyword ("from", Token.FROM);
+ AddKeyword ("join", Token.JOIN);
+ AddKeyword ("on", Token.ON);
+ AddKeyword ("equals", Token.EQUALS);
+ AddKeyword ("select", Token.SELECT);
+ AddKeyword ("group", Token.GROUP);
+ AddKeyword ("by", Token.BY);
+ AddKeyword ("let", Token.LET);
+ AddKeyword ("orderby", Token.ORDERBY);
+ AddKeyword ("ascending", Token.ASCENDING);
+ AddKeyword ("descending", Token.DESCENDING);
+ AddKeyword ("into", Token.INTO);
}
//
// Class initializer
//
static Tokenizer ()
+ {
+ Reset ();
+ }
+
+ public static void Reset ()
{
InitTokens ();
csharp_format_info = NumberFormatInfo.InvariantInfo;
styles = NumberStyles.Float;
-
+
string_builder = new System.Text.StringBuilder ();
}
int res = (int) o;
- if (handle_get_set == false && (res == Token.GET || res == Token.SET))
- return -1;
- if (handle_remove_add == false && (res == Token.REMOVE || res == Token.ADD))
+ if (!handle_get_set && (res == Token.GET || res == Token.SET))
return -1;
- if (handle_assembly == false && res == Token.ASSEMBLY)
+ if (!handle_remove_add && (res == Token.REMOVE || res == Token.ADD))
return -1;
-#if GMCS_SOURCE
- if (handle_constraints == false && res == Token.WHERE)
+ if (parsing_declaration == 0 && res == Token.EXTERN)
+ return Token.EXTERN_ALIAS;
+
+ //
+ // A query expression is any expression that starts with `from identifier'
+ // followed by any token except ; , =
+ //
+ if (query_parsing == 0) {
+ if (res == Token.FROM && !lambda_arguments_parsing) {
+ PushPosition ();
+ // HACK: to disable generics micro-parser, because PushPosition does not
+ // store identifiers array
+ parsing_generic_less_than = 1;
+ switch (xtoken ()) {
+ case Token.IDENTIFIER:
+ case Token.INT:
+ case Token.BOOL:
+ case Token.BYTE:
+ case Token.CHAR:
+ case Token.DECIMAL:
+ case Token.FLOAT:
+ case Token.LONG:
+ case Token.OBJECT:
+ case Token.STRING:
+ case Token.UINT:
+ case Token.ULONG:
+ int next_token = xtoken ();
+ if (next_token == Token.SEMICOLON || next_token == Token.COMMA || next_token == Token.EQUALS)
+ goto default;
+
+ ++query_parsing;
+ if (RootContext.Version <= LanguageVersion.ISO_2)
+ Report.FeatureIsNotAvailable (Location, "query expressions");
+ break;
+ case Token.VOID:
+ Expression.Error_VoidInvalidInTheContext (Location);
+ break;
+ default:
+ PopPosition ();
+ // HACK: A token is not a keyword so we need to restore identifiers buffer
+ // which has been overwritten before we grabbed the identifier
+ id_builder [0] = 'f'; id_builder [1] = 'r'; id_builder [2] = 'o'; id_builder [3] = 'm';
+ return -1;
+ }
+ PopPosition ();
+ return res;
+ }
+
+ if (res > Token.QUERY_FIRST_TOKEN && res < Token.QUERY_LAST_TOKEN)
+ return -1;
+ }
+
+ if (res == Token.WHERE && !handle_where && query_parsing == 0)
return -1;
-#endif
- return res;
+ return res;
}
public Location Location {
get { return current_location; }
}
- void define (string def)
- {
- if (!RootContext.AllDefines.Contains (def)){
- RootContext.AllDefines [def] = true;
- }
- if (defines.Contains (def))
- return;
- defines [def] = true;
- }
-
- public Tokenizer (SeekableStreamReader input, SourceFile file, ArrayList defs)
+ public Tokenizer (SeekableStreamReader input, CompilationUnit file)
{
this.ref_name = file;
this.file_name = file;
- linq = RootContext.Version == LanguageVersion.LINQ;
reader = input;
putback_char = -1;
- if (defs != null){
- defines = new Hashtable ();
- foreach (string def in defs)
- define (def);
- }
-
xml_comment_buffer = new StringBuilder ();
//
// FIXME: This could be `Location.Push' but we have to
// find out why the MS compiler allows this
//
- Mono.CSharp.Location.Push (file);
+ Mono.CSharp.Location.Push (file, file);
}
static bool is_identifier_start_character (char c)
public static bool IsKeyword (string s)
{
- return keywordStrings [s] != null;
+ return keyword_strings [s] != null;
+ }
+
+ //
+ // Open parens micro parser. Detects both lambda and cast ambiguity.
+ //
+
+ int TokenizeOpenParens ()
+ {
+ int ntoken = -1;
+ int ptoken;
+ bool cast_posible = false;
+ bool cast_not = false;
+ bool has_star = false;
+
+ while (true) {
+ ptoken = ntoken;
+ ntoken = xtoken ();
+
+ switch (ntoken) {
+ case Token.CLOSE_PARENS:
+ ntoken = xtoken ();
+
+ //
+ // Token is a lambda
+ //
+ if (ntoken == Token.ARROW) {
+ if (RootContext.Version <= LanguageVersion.ISO_2)
+ Report.FeatureIsNotAvailable (Location, "lambda expressions");
+
+ return Token.OPEN_PARENS_LAMBDA;
+ }
+
+ //
+ // Token is possible cast, parser will decide later
+ //
+ if (cast_posible && !cast_not) {
+ switch (ntoken) {
+ //
+ // Indirection is a special
+ // cast: (int*)&a;
+ // binary: (C)&a;
+ //
+ case Token.BITWISE_AND:
+ case Token.OP_INC:
+ case Token.OP_DEC:
+ case Token.PLUS:
+ case Token.MINUS:
+ if (!has_star)
+ break;
+ return Token.OPEN_PARENS_CAST;
+
+ case Token.OPEN_PARENS:
+ case Token.BANG:
+ case Token.TILDE:
+ case Token.IDENTIFIER:
+ case Token.LITERAL_INTEGER:
+ case Token.LITERAL_FLOAT:
+ case Token.LITERAL_DOUBLE:
+ case Token.LITERAL_DECIMAL:
+ case Token.LITERAL_CHARACTER:
+ case Token.LITERAL_STRING:
+ case Token.BASE:
+ case Token.CHECKED:
+ case Token.DELEGATE:
+ case Token.FALSE:
+ case Token.FIXED:
+ case Token.NEW:
+ case Token.NULL:
+ case Token.SIZEOF:
+ case Token.THIS:
+ case Token.THROW:
+ case Token.TRUE:
+ case Token.TYPEOF:
+ case Token.UNCHECKED:
+ case Token.UNSAFE:
+ case Token.DEFAULT:
+
+ //
+ // These can be part of a member access
+ //
+ case Token.INT:
+ case Token.UINT:
+ case Token.SHORT:
+ case Token.USHORT:
+ case Token.LONG:
+ case Token.ULONG:
+ case Token.DOUBLE:
+ case Token.FLOAT:
+ case Token.CHAR:
+ case Token.BYTE:
+ case Token.DECIMAL:
+ return Token.OPEN_PARENS_CAST;
+ }
+ }
+ return Token.OPEN_PARENS;
+
+ case Token.DOT:
+ if (ptoken != Token.IDENTIFIER && ptoken != Token.OP_GENERICS_GT)
+ goto default;
+ continue;
+
+ case Token.IDENTIFIER:
+ switch (ptoken) {
+ case Token.DOT:
+ case Token.OP_GENERICS_LT:
+ case Token.COMMA:
+ case Token.DOUBLE_COLON:
+ case -1:
+ cast_posible = true;
+ continue;
+ default:
+ cast_not = true;
+ continue;
+ }
+
+ case Token.STAR:
+ has_star = true;
+ continue;
+
+ case Token.OBJECT:
+ case Token.STRING:
+ case Token.BOOL:
+ case Token.DECIMAL:
+ case Token.FLOAT:
+ case Token.DOUBLE:
+ case Token.SBYTE:
+ case Token.BYTE:
+ case Token.SHORT:
+ case Token.USHORT:
+ case Token.INT:
+ case Token.UINT:
+ case Token.LONG:
+ case Token.ULONG:
+ case Token.CHAR:
+ case Token.VOID:
+ case Token.OP_GENERICS_GT:
+ case Token.OP_GENERICS_LT:
+ case Token.INTERR_NULLABLE:
+ case Token.OPEN_BRACKET:
+ case Token.CLOSE_BRACKET:
+ case Token.COMMA:
+ case Token.DOUBLE_COLON:
+ cast_posible = true;
+ continue;
+
+ case Token.REF:
+ case Token.OUT:
+ cast_not = true;
+ continue;
+
+ default:
+ return Token.OPEN_PARENS;
+ }
+ }
}
public static bool IsValidIdentifier (string s)
return true;
}
- bool parse_generic_dimension (out int dimension)
- {
- dimension = 1;
-
- again:
- int the_token = token ();
- if (the_token == Token.OP_GENERICS_GT)
- return true;
- else if (the_token == Token.COMMA) {
- dimension++;
- goto again;
- }
-
- return false;
- }
-
bool parse_less_than ()
{
start:
case Token.VOID:
break;
+ case Token.OP_GENERICS_GT:
+ return true;
+
default:
return false;
}
return true;
else if (the_token == Token.COMMA || the_token == Token.DOT || the_token == Token.DOUBLE_COLON)
goto start;
- else if (the_token == Token.INTERR || the_token == Token.STAR)
+ else if (the_token == Token.INTERR_NULLABLE || the_token == Token.STAR)
goto again;
else if (the_token == Token.OP_GENERICS_LT) {
if (!parse_less_than ())
return false;
}
-#if GMCS_SOURCE
- public void PutbackNullable ()
- {
- if (nullable_pos < 0)
- throw new Exception ();
-
- current_token = -1;
- val = null;
- reader.Position = nullable_pos;
-
- putback_char = '?';
- }
-
- public void PutbackCloseParens ()
+ bool parse_generic_dimension (out int dimension)
{
- putback_char = ')';
- }
-
+ dimension = 1;
- int nullable_pos = -1;
+ again:
+ int the_token = token ();
+ if (the_token == Token.OP_GENERICS_GT)
+ return true;
+ else if (the_token == Token.COMMA) {
+ dimension++;
+ goto again;
+ }
- public void CheckNullable (bool is_nullable)
- {
- if (is_nullable)
- nullable_pos = reader.Position;
- else
- nullable_pos = -1;
+ return false;
}
-#endif
public int peek_token ()
{
return the_token;
}
-
- bool parse_namespace_or_typename (int next)
- {
- if (next == -1)
- next = peek_token ();
- while (next == Token.IDENTIFIER){
- token ();
- again:
- next = peek_token ();
- if (next == Token.DOT || next == Token.DOUBLE_COLON){
- token ();
- next = peek_token ();
- continue;
- }
- if (next == Token.OP_GENERICS_LT){
- token ();
- if (!parse_less_than ())
- throw new InternalErrorException ("cannot happen");
- goto again;
- }
- return true;
- }
-
- return false;
- }
-
- bool is_simple_type (int token)
- {
- return (token == Token.BOOL ||
- token == Token.DECIMAL ||
- token == Token.SBYTE ||
- token == Token.BYTE ||
- token == Token.SHORT ||
- token == Token.USHORT ||
- token == Token.INT ||
- token == Token.UINT ||
- token == Token.LONG ||
- token == Token.ULONG ||
- token == Token.CHAR ||
- token == Token.FLOAT ||
- token == Token.DOUBLE);
- }
-
- bool is_builtin_reference_type (int token)
- {
- return (token == Token.OBJECT || token == Token.STRING);
- }
-
- bool parse_opt_rank (int next)
- {
- while (true){
- if (next != Token.OPEN_BRACKET)
- return true;
-
- token ();
- while (true){
- next = token ();
- if (next == Token.CLOSE_BRACKET){
- next = peek_token ();
- break;
- }
- if (next == Token.COMMA)
- continue;
- return false;
- }
- }
- }
-
- bool parse_type ()
- {
- int next = peek_token ();
-
- if (is_simple_type (next)){
- token ();
- next = peek_token ();
- if (next == Token.INTERR)
- token ();
- return parse_opt_rank (peek_token ());
- }
- if (parse_namespace_or_typename (next)){
- next = peek_token ();
- if (next == Token.INTERR)
- token ();
- return parse_opt_rank (peek_token ());
- } else if (is_builtin_reference_type (next)){
- token ();
- return parse_opt_rank (peek_token ());
- }
-
- return false;
- }
-
- //
- // Invoked after '(' has been seen and tries to parse
- // a type expression followed by an identifier, if this
- // is the case, instead of returning an OPEN_PARENS token
- // we return a special token that triggers lambda parsing.
- //
- // This is needed because we can not introduce the
- // explicitly_typed_lambda_parameter_list after a '(' in the
- // grammar without introducing reduce/reduce conflicts.
- //
- // We need to parse a type and if it is followed by an
- // identifier, we know it has to be parsed as a lambda
- // expression.
- //
- // the type expression can be prefixed with `ref' or `out'
- //
- public bool parse_type_and_parameter ()
- {
- int next = peek_token ();
-
- if (next == Token.REF || next == Token.OUT)
- token ();
-
- if (parse_type ()){
- next = peek_token ();
- if (next == Token.IDENTIFIER)
- return true;
- }
- return false;
- }
-
int parsing_generic_less_than = 0;
int is_punct (char c, ref bool doread)
case ']':
return Token.CLOSE_BRACKET;
case '(':
- if (linq){
+ val = Location;
+ //
+ // An expression versions of parens can appear in block context only
+ //
+ if (parsing_block != 0 && !lambda_arguments_parsing) {
+ lambda_arguments_parsing = true;
PushPosition ();
- bool is_type_and_parameter = parse_type_and_parameter ();
+ t = TokenizeOpenParens ();
PopPosition ();
-
- if (is_type_and_parameter)
- return Token.OPEN_PARENS_LAMBDA;
- else
- return Token.OPEN_PARENS;
- } else
- return Token.OPEN_PARENS;
- case ')': {
- if (deambiguate_close_parens == 0)
- return Token.CLOSE_PARENS;
-
- --deambiguate_close_parens;
-
- PushPosition ();
-
- int new_token = xtoken ();
-
- PopPosition ();
-
- if (new_token == Token.OPEN_PARENS)
- return Token.CLOSE_PARENS_OPEN_PARENS;
- else if (new_token == Token.MINUS)
- return Token.CLOSE_PARENS_MINUS;
- else if (IsCastToken (new_token))
- return Token.CLOSE_PARENS_CAST;
- else
- return Token.CLOSE_PARENS_NO_CAST;
- }
+ lambda_arguments_parsing = false;
+ return t;
+ }
+ return Token.OPEN_PARENS;
+ case ')':
+ return Token.CLOSE_PARENS;
case ',':
return Token.COMMA;
case ';':
val = Location;
return Token.TILDE;
case '?':
- return Token.INTERR;
+ return TokenizePossibleNullableType ();
}
-#if GMCS_SOURCE
+
if (c == '<') {
if (parsing_generic_less_than++ > 0)
return Token.OP_GENERICS_LT;
// Save current position and parse next token.
PushPosition ();
bool is_generic_lt = parse_less_than ();
- PopPosition ();
-
if (is_generic_lt) {
- return Token.OP_GENERICS_LT;
- } else
- parsing_generic_less_than = 0;
+ int rt;
+ if (parsing_generic_declaration && token () != Token.DOT) {
+ rt = Token.OP_GENERICS_LT_DECL;
+ } else {
+ rt = Token.OP_GENERICS_LT;
+ }
+ PopPosition ();
+ return rt;
+ }
+
+ PopPosition ();
+ parsing_generic_less_than = 0;
- d = peekChar ();
+ d = peek_char ();
if (d == '<'){
- getChar ();
- d = peekChar ();
+ get_char ();
+ d = peek_char ();
if (d == '='){
doread = true;
}
return Token.OP_LT;
} else if (c == '>') {
- if (parsing_generic_less_than > 0) {
+ d = peek_char ();
+
+ if (d == '='){
+ doread = true;
+ return Token.OP_GE;
+ }
+
+ if (parsing_generic_less_than > 1 || (parsing_generic_less_than == 1 && d != '>')) {
parsing_generic_less_than--;
return Token.OP_GENERICS_GT;
}
- d = peekChar ();
- if (d == '>'){
- getChar ();
- d = peekChar ();
+ if (d == '>') {
+ get_char ();
+ d = peek_char ();
- if (d == '='){
+ if (d == '=') {
doread = true;
return Token.OP_SHIFT_RIGHT_ASSIGN;
}
return Token.OP_SHIFT_RIGHT;
- } else if (d == '='){
- doread = true;
- return Token.OP_GE;
}
+
return Token.OP_GT;
}
-#endif
- d = peekChar ();
+
+ d = peek_char ();
if (c == '+'){
if (d == '+') {
val = Location;
return Token.ARROW;
}
+
return Token.ASSIGN;
}
if (c == '^'){
if (d == '='){
doread = true;
- return Token.OP_XOR_ASSIGN;
- }
- return Token.CARRET;
- }
-
-#if !GMCS_SOURCE
- if (c == '<'){
- if (d == '<'){
- getChar ();
- d = peekChar ();
-
- if (d == '='){
- doread = true;
- return Token.OP_SHIFT_LEFT_ASSIGN;
- }
- return Token.OP_SHIFT_LEFT;
- } else if (d == '='){
- doread = true;
- return Token.OP_LE;
- }
- return Token.OP_LT;
- }
-
- if (c == '>'){
- if (d == '>'){
- getChar ();
- d = peekChar ();
-
- if (d == '='){
- doread = true;
- return Token.OP_SHIFT_RIGHT_ASSIGN;
- }
- return Token.OP_SHIFT_RIGHT;
- } else if (d == '='){
- doread = true;
- return Token.OP_GE;
+ return Token.OP_XOR_ASSIGN;
}
- return Token.OP_GT;
+ return Token.CARRET;
}
-#endif
+
if (c == ':'){
if (d == ':'){
doread = true;
return Token.ERROR;
}
- int deambiguate_close_parens = 0;
-
- public void Deambiguate_CloseParens (object expression)
+ //
+ // Tonizes `?' using custom disambiguous rules to return one
+ // of following tokens: INTERR_NULLABLE, OP_COALESCING, INTERR
+ //
+ // Tricky expression look like:
+ //
+ // Foo ? a = x ? b : c;
+ //
+ int TokenizePossibleNullableType ()
{
- putback (')');
+ if (parsing_block == 0)
+ return Token.INTERR_NULLABLE;
- // When any binary operation is used we are sure it is not a cast
- if (expression is Binary)
- return;
+ int d = peek_char ();
+ if (d == '?') {
+ get_char ();
+ return Token.OP_COALESCING;
+ }
+
+ switch (current_token) {
+ case Token.CLOSE_PARENS:
+ case Token.TRUE:
+ case Token.FALSE:
+ case Token.NULL:
+ case Token.LITERAL_INTEGER:
+ case Token.LITERAL_STRING:
+ return Token.INTERR;
+ }
+
+ if (d != ' ') {
+ if (d == ',' || d == ';' || d == '>')
+ return Token.INTERR_NULLABLE;
+ if (d == '*' || (d >= '0' && d <= '9'))
+ return Token.INTERR;
+ }
+
+ PushPosition ();
+ int next_token;
+ switch (xtoken ()) {
+ case Token.LITERAL_INTEGER:
+ case Token.LITERAL_STRING:
+ case Token.LITERAL_CHARACTER:
+ case Token.LITERAL_DECIMAL:
+ case Token.LITERAL_DOUBLE:
+ case Token.LITERAL_FLOAT:
+ case Token.TRUE:
+ case Token.FALSE:
+ case Token.NULL:
+ case Token.THIS:
+ next_token = Token.INTERR;
+ break;
+
+ case Token.SEMICOLON:
+ case Token.COMMA:
+ case Token.CLOSE_PARENS:
+ case Token.OPEN_BRACKET:
+ case Token.OP_GENERICS_GT:
+ next_token = Token.INTERR_NULLABLE;
+ break;
+
+ default:
+ next_token = -1;
+ break;
+ }
- deambiguate_close_parens++;
+ if (next_token == -1) {
+ switch (xtoken ()) {
+ case Token.COMMA:
+ case Token.SEMICOLON:
+ case Token.OPEN_BRACE:
+ case Token.CLOSE_PARENS:
+ case Token.IN:
+ next_token = Token.INTERR_NULLABLE;
+ break;
+
+ case Token.COLON:
+ next_token = Token.INTERR;
+ break;
+
+ default:
+ int ntoken;
+ int interrs = 1;
+ int colons = 0;
+ //
+ // All shorcuts failed, do it hard way
+ //
+ while ((ntoken = xtoken ()) != Token.EOF) {
+ if (ntoken == Token.SEMICOLON)
+ break;
+
+ if (ntoken == Token.COLON) {
+ if (++colons == interrs)
+ break;
+ continue;
+ }
+
+ if (ntoken == Token.INTERR) {
+ ++interrs;
+ continue;
+ }
+ }
+
+ next_token = colons != interrs ? Token.INTERR_NULLABLE : Token.INTERR;
+ break;
+ }
+ }
+
+ PopPosition ();
+ return next_token;
}
bool decimal_digits (int c)
}
//
- // We use peekChar2, because decimal_digits needs to do a
+ // We use peek_char2, because decimal_digits needs to do a
// 2-character look-ahead (5.ToString for example).
//
- while ((d = peekChar2 ()) != -1){
+ while ((d = peek_char2 ()) != -1){
if (d >= '0' && d <= '9'){
if (number_pos == max_number_size)
Error_NumericConstantTooLong ();
number_builder [number_pos++] = (char) d;
- getChar ();
+ get_char ();
seen_digits = true;
} else
break;
{
return (e >= '0' && e <= '9') || (e >= 'A' && e <= 'F') || (e >= 'a' && e <= 'f');
}
-
+
static int real_type_suffix (int c)
{
int t;
if (is_unsigned)
scanning = false;
is_unsigned = true;
- getChar ();
+ get_char ();
break;
case 'l':
- if (!is_unsigned && (RootContext.WarningLevel >= 4)){
+ if (!is_unsigned){
//
// if we have not seen anything in between
// report this error
if (is_long)
scanning = false;
is_long = true;
- getChar ();
+ get_char ();
break;
case 'L':
if (is_long)
scanning = false;
is_long = true;
- getChar ();
+ get_char ();
break;
default:
scanning = false;
break;
}
- c = peekChar ();
+ c = peek_char ();
} while (scanning);
}
int d;
ulong ul;
- getChar ();
- while ((d = peekChar ()) != -1){
+ get_char ();
+ while ((d = peek_char ()) != -1){
if (is_hex (d)){
number_builder [number_pos++] = (char) d;
- getChar ();
+ get_char ();
} else
break;
}
return Token.LITERAL_INTEGER;
}
- return integer_type_suffix (ul, peekChar ());
+ return integer_type_suffix (ul, peek_char ());
}
//
if (c >= '0' && c <= '9'){
if (c == '0'){
- int peek = peekChar ();
+ int peek = peek_char ();
if (peek == 'x' || peek == 'X')
return handle_hex ();
}
decimal_digits (c);
- c = getChar ();
+ c = get_char ();
}
//
if (c == '.'){
if (decimal_digits ('.')){
is_real = true;
- c = getChar ();
+ c = get_char ();
} else {
putback ('.');
number_pos--;
if (number_pos == max_number_size)
Error_NumericConstantTooLong ();
number_builder [number_pos++] = 'e';
- c = getChar ();
+ c = get_char ();
if (c == '+'){
if (number_pos == max_number_size)
}
decimal_digits (c);
- c = getChar ();
+ c = get_char ();
}
type = real_type_suffix (c);
//
// Accepts exactly count (4 or 8) hex, no more no less
//
- int getHex (int count, out bool error)
+ int getHex (int count, out int surrogate, out bool error)
{
int i;
int total = 0;
int c;
int top = count != -1 ? count : 4;
- getChar ();
+ get_char ();
error = false;
+ surrogate = 0;
for (i = 0; i < top; i++){
- c = getChar ();
-
+ c = get_char ();
+
if (c >= '0' && c <= '9')
c = (int) c - (int) '0';
else if (c >= 'A' && c <= 'F')
total = (total * 16) + c;
if (count == -1){
- int p = peekChar ();
+ int p = peek_char ();
if (p == -1)
break;
if (!is_hex ((char)p))
break;
}
}
+
+ if (top == 8) {
+ if (total > 0x0010FFFF) {
+ error = true;
+ return 0;
+ }
+
+ if (total >= 0x00010000) {
+ surrogate = ((total - 0x00010000) % 0x0400 + 0xDC00);
+ total = ((total - 0x00010000) / 0x0400 + 0xD800);
+ }
+ }
+
return total;
}
- int escape (int c)
+ int escape (int c, out int surrogate)
{
bool error;
int d;
int v;
- d = peekChar ();
- if (c != '\\')
+ d = peek_char ();
+ if (c != '\\') {
+ surrogate = 0;
return c;
+ }
switch (d){
case 'a':
case '\'':
v = '\''; break;
case 'x':
- v = getHex (-1, out error);
+ v = getHex (-1, out surrogate, out error);
if (error)
goto default;
return v;
case 'u':
- v = getHex (4, out error);
- if (error)
- goto default;
- return v;
case 'U':
- v = getHex (8, out error);
- if (error)
- goto default;
- return v;
+ return EscapeUnicode (d, out surrogate);
default:
+ surrogate = 0;
Report.Error (1009, Location, "Unrecognized escape sequence `\\{0}'", ((char)d).ToString ());
return d;
}
- getChar ();
+
+ get_char ();
+ surrogate = 0;
return v;
}
- int getChar ()
+ int EscapeUnicode (int ch, out int surrogate)
+ {
+ bool error;
+ if (ch == 'U') {
+ ch = getHex (8, out surrogate, out error);
+ } else {
+ ch = getHex (4, out surrogate, out error);
+ }
+
+ if (error)
+ Report.Error (1009, Location, "Unrecognized escape sequence");
+
+ return ch;
+ }
+
+ int get_char ()
{
int x;
if (putback_char != -1) {
} else
x = reader.Read ();
if (x == '\n') {
- line++;
- ref_line++;
- previous_col = col;
- col = 0;
- }
- else
+ advance_line ();
+ } else {
col++;
+ }
return x;
}
- int peekChar ()
+ void advance_line ()
+ {
+ line++;
+ ref_line++;
+ previous_col = col;
+ col = 0;
+ }
+
+ int peek_char ()
{
if (putback_char != -1)
return putback_char;
return putback_char;
}
- int peekChar2 ()
+ int peek_char2 ()
{
if (putback_char != -1)
return putback_char;
public bool advance ()
{
- return peekChar () != -1;
+ return peek_char () != -1;
}
public Object Value {
return val;
}
- static bool IsCastToken (int token)
- {
- switch (token) {
- case Token.BANG:
- case Token.TILDE:
- case Token.IDENTIFIER:
- case Token.LITERAL_INTEGER:
- case Token.LITERAL_FLOAT:
- case Token.LITERAL_DOUBLE:
- case Token.LITERAL_DECIMAL:
- case Token.LITERAL_CHARACTER:
- case Token.LITERAL_STRING:
- case Token.BASE:
- case Token.CHECKED:
- case Token.DELEGATE:
- case Token.FALSE:
- case Token.FIXED:
- case Token.NEW:
- case Token.NULL:
- case Token.SIZEOF:
- case Token.THIS:
- case Token.THROW:
- case Token.TRUE:
- case Token.TYPEOF:
- case Token.UNCHECKED:
- case Token.UNSAFE:
-#if GMCS_SOURCE
- case Token.DEFAULT:
-#endif
-
- //
- // These can be part of a member access
- //
- case Token.INT:
- case Token.UINT:
- case Token.SHORT:
- case Token.USHORT:
- case Token.LONG:
- case Token.ULONG:
- case Token.DOUBLE:
- case Token.FLOAT:
- case Token.CHAR:
- return true;
-
- default:
- return false;
- }
- }
-
public int token ()
{
current_token = xtoken ();
-#if GMCS_SOURCE
if (current_token != Token.DEFAULT)
return current_token;
int c = xtoken();
if (c == -1)
current_token = Token.ERROR;
- else if (c == Token.OPEN_PARENS)
- current_token = Token.DEFAULT_OPEN_PARENS;
else if (c == Token.COLON)
current_token = Token.DEFAULT_COLON;
else
PopPosition();
-#endif
+
return current_token;
}
static StringBuilder static_cmd_arg = new System.Text.StringBuilder ();
-
+
void get_cmd_arg (out string cmd, out string arg)
{
int c;
tokens_seen = false;
arg = "";
- static_cmd_arg.Length = 0;
// skip over white space
- while ((c = getChar ()) != -1 && (c != '\n') && ((c == '\r') || (c == ' ') || (c == '\t')))
- ;
-
- while ((c != -1) && (c != '\n') && (c != ' ') && (c != '\t') && (c != '\r')){
- if (is_identifier_part_character ((char) c)){
- static_cmd_arg.Append ((char) c);
- c = getChar ();
- } else {
- putback (c);
- break;
+ do {
+ c = get_char ();
+ } while (c == '\r' || c == ' ' || c == '\t');
+
+ static_cmd_arg.Length = 0;
+ while (c != -1 && is_identifier_part_character ((char)c)) {
+ static_cmd_arg.Append ((char)c);
+ c = get_char ();
+ if (c == '\\') {
+ int peek = peek_char ();
+ if (peek == 'U' || peek == 'u') {
+ int surrogate;
+ c = EscapeUnicode (c, out surrogate);
+ if (surrogate != 0) {
+ if (is_identifier_part_character ((char) c))
+ static_cmd_arg.Append ((char) c);
+ c = surrogate;
+ }
+ }
}
}
cmd = static_cmd_arg.ToString ();
- if (c == '\n' || c == '\r'){
- return;
- }
-
// skip over white space
- while ((c = getChar ()) != -1 && (c != '\n') && ((c == '\r') || (c == ' ') || (c == '\t')))
- ;
+ while (c == '\r' || c == ' ' || c == '\t')
+ c = get_char ();
- if (c == '\n'){
- return;
- } else if (c == '\r'){
- return;
- } else if (c == -1){
- arg = "";
- return;
- }
-
static_cmd_arg.Length = 0;
- static_cmd_arg.Append ((char) c);
-
- while ((c = getChar ()) != -1 && (c != '\n') && (c != '\r')){
+ while (c != -1 && c != '\n' && c != '\r') {
+ if (c == '\\') {
+ int peek = peek_char ();
+ if (peek == 'U' || peek == 'u') {
+ int surrogate;
+ c = EscapeUnicode (c, out surrogate);
+ if (surrogate != 0) {
+ if (is_identifier_part_character ((char) c))
+ static_cmd_arg.Append ((char) c);
+ c = surrogate;
+ }
+ }
+ }
static_cmd_arg.Append ((char) c);
+ c = get_char ();
}
- arg = static_cmd_arg.ToString ();
+ if (static_cmd_arg.Length != 0)
+ arg = static_cmd_arg.ToString ();
}
//
if (arg == "default"){
ref_line = line;
ref_name = file_name;
- Location.Push (ref_name);
+ hidden = false;
+ Location.Push (file_name, ref_name);
return true;
} else if (arg == "hidden"){
- //
- // We ignore #line hidden
- //
+ hidden = true;
return true;
}
char [] quotes = { '\"' };
string name = arg.Substring (pos). Trim (quotes);
- ref_name = Location.LookupFile (name);
- file_name.HasLineDirective = true;
- ref_name.HasLineDirective = true;
- Location.Push (ref_name);
+ ref_name = Location.LookupFile (file_name, name);
+ file_name.AddFile (ref_name);
+ hidden = false;
+ Location.Push (file_name, ref_name);
} else {
ref_line = System.Int32.Parse (arg);
+ hidden = false;
}
} catch {
return false;
//
// Handles #define and #undef
//
- void PreProcessDefinition (bool is_define, string arg, bool caller_is_taking)
+ void PreProcessDefinition (bool is_define, string ident, bool caller_is_taking)
{
- if (arg.Length == 0 || arg == "true" || arg == "false"){
- Report.Error (1001, Location, "Missing identifer to pre-processor directive");
+ if (ident.Length == 0 || ident == "true" || ident == "false"){
+ Report.Error (1001, Location, "Missing identifier to pre-processor directive");
return;
}
- if (arg.IndexOfAny (simple_whitespaces) != -1){
+ if (ident.IndexOfAny (simple_whitespaces) != -1){
Error_EndLineExpected ();
return;
}
- if (!is_identifier_start_character (arg [0]))
- Report.Error (1001, Location, "Identifier expected: " + arg);
+ if (!is_identifier_start_character (ident [0]))
+ Report.Error (1001, Location, "Identifier expected: {0}", ident);
- foreach (char c in arg.Substring (1)){
+ foreach (char c in ident.Substring (1)){
if (!is_identifier_part_character (c)){
- Report.Error (1001, Location, "Identifier expected: " + arg);
+ Report.Error (1001, Location, "Identifier expected: {0}", ident);
return;
}
}
if (!caller_is_taking)
return;
- if (is_define){
- if (defines == null)
- defines = new Hashtable ();
- define (arg);
- } else {
- if (defines == null)
+ if (is_define) {
+ //
+ // #define ident
+ //
+ if (RootContext.IsConditionalDefined (ident))
return;
- if (defines.Contains (arg))
- defines.Remove (arg);
+
+ file_name.AddDefine (ident);
+ } else {
+ //
+ // #undef ident
+ //
+ file_name.AddUndefine (ident);
+ }
+ }
+
+ static byte read_hex (string arg, int pos, out bool error)
+ {
+ error = false;
+
+ int total;
+ char c = arg [pos];
+
+ if ((c >= '0') && (c <= '9'))
+ total = (int) c - (int) '0';
+ else if ((c >= 'A') && (c <= 'F'))
+ total = (int) c - (int) 'A' + 10;
+ else if ((c >= 'a') && (c <= 'f'))
+ total = (int) c - (int) 'a' + 10;
+ else {
+ error = true;
+ return 0;
+ }
+
+ total *= 16;
+ c = arg [pos+1];
+
+ if ((c >= '0') && (c <= '9'))
+ total += (int) c - (int) '0';
+ else if ((c >= 'A') && (c <= 'F'))
+ total += (int) c - (int) 'A' + 10;
+ else if ((c >= 'a') && (c <= 'f'))
+ total += (int) c - (int) 'a' + 10;
+ else {
+ error = true;
+ return 0;
+ }
+
+ return (byte) total;
+ }
+
+ /// <summary>
+ /// Handles #pragma checksum
+ /// </summary>
+ bool PreProcessPragmaChecksum (string arg)
+ {
+ if ((arg [0] != ' ') && (arg [0] != '\t'))
+ return false;
+
+ arg = arg.Trim (simple_whitespaces);
+ if ((arg.Length < 2) || (arg [0] != '"'))
+ return false;
+
+ StringBuilder file_sb = new StringBuilder ();
+
+ int pos = 1;
+ char ch;
+ while ((ch = arg [pos++]) != '"') {
+ if (pos >= arg.Length)
+ return false;
+
+ if (ch == '\\') {
+ if (pos+1 >= arg.Length)
+ return false;
+ ch = arg [pos++];
+ }
+
+ file_sb.Append (ch);
+ }
+
+ if ((pos+2 >= arg.Length) || ((arg [pos] != ' ') && (arg [pos] != '\t')))
+ return false;
+
+ arg = arg.Substring (pos).Trim (simple_whitespaces);
+ if ((arg.Length < 42) || (arg [0] != '"') || (arg [1] != '{') ||
+ (arg [10] != '-') || (arg [15] != '-') || (arg [20] != '-') ||
+ (arg [25] != '-') || (arg [38] != '}') || (arg [39] != '"'))
+ return false;
+
+ bool error;
+ byte[] guid_bytes = new byte [16];
+
+ for (int i = 0; i < 4; i++) {
+ guid_bytes [i] = read_hex (arg, 2+2*i, out error);
+ if (error)
+ return false;
+ }
+ for (int i = 0; i < 2; i++) {
+ guid_bytes [i+4] = read_hex (arg, 11+2*i, out error);
+ if (error)
+ return false;
+ guid_bytes [i+6] = read_hex (arg, 16+2*i, out error);
+ if (error)
+ return false;
+ guid_bytes [i+8] = read_hex (arg, 21+2*i, out error);
+ if (error)
+ return false;
+ }
+
+ for (int i = 0; i < 6; i++) {
+ guid_bytes [i+10] = read_hex (arg, 26+2*i, out error);
+ if (error)
+ return false;
}
+
+ arg = arg.Substring (40).Trim (simple_whitespaces);
+ if ((arg.Length < 34) || (arg [0] != '"') || (arg [33] != '"'))
+ return false;
+
+ byte[] checksum_bytes = new byte [16];
+ for (int i = 0; i < 16; i++) {
+ checksum_bytes [i] = read_hex (arg, 1+2*i, out error);
+ if (error)
+ return false;
+ }
+
+ arg = arg.Substring (34).Trim (simple_whitespaces);
+ if (arg.Length > 0)
+ return false;
+
+ SourceFile file = Location.LookupFile (file_name, file_sb.ToString ());
+ file.SetChecksum (guid_bytes, checksum_bytes);
+ ref_name.AutoGenerated = true;
+ return true;
}
/// <summary>
const string warning = "warning";
const string w_disable = "warning disable";
const string w_restore = "warning restore";
+ const string checksum = "checksum";
if (arg == w_disable) {
- Report.RegisterWarningRegion (Location).WarningDisable (line);
+ Report.RegisterWarningRegion (Location).WarningDisable (Location.Row);
return;
}
if (arg == w_restore) {
- Report.RegisterWarningRegion (Location).WarningEnable (line);
+ Report.RegisterWarningRegion (Location).WarningEnable (Location.Row);
return;
}
return;
}
+ if (arg.StartsWith (checksum)) {
+ if (!PreProcessPragmaChecksum (arg.Substring (checksum.Length)))
+ Warning_InvalidPragmaChecksum ();
+ return;
+ }
+
Report.Warning (1633, 1, Location, "Unrecognized #pragma directive");
}
return true;
if (s == "false")
return false;
-
- if (defines == null)
- return false;
- if (defines.Contains (s))
- return true;
- return false;
+ return file_name.IsConditionalDefined (s);
}
bool pp_primary (ref string s)
{
Report.Error (
1028, Location,
- "Unexpected processor directive (" + extra + ")");
+ "Unexpected processor directive ({0})", extra);
}
void Error_TokensSeen ()
Report.Error (1025, Location, "Single-line comment or end-of-line expected");
}
+ void Warning_InvalidPragmaChecksum ()
+ {
+ Report.Warning (1695, 1, Location,
+ "Invalid #pragma checksum syntax; should be " +
+ "#pragma checksum \"filename\" " +
+ "\"{XXXXXXXX-XXXX-XXXX-XXXX-XXXXXXXXXXXX}\" \"XXXX...\"");
+ }
//
// if true, then the code continues processing the code
// if false, the code stays in a loop until another directive is
switch (cmd){
case "error":
- Report.Error (1029, Location, "#error: '" + arg + "'");
+ Report.Error (1029, Location, "#error: '{0}'", arg);
return true;
case "warning":
case "pragma":
if (RootContext.Version == LanguageVersion.ISO_1) {
- Report.FeatureIsNotISO1 (Location, "#pragma");
+ Report.FeatureIsNotAvailable (Location, "#pragma");
return true;
}
{
int c;
string_builder.Length = 0;
-
- while ((c = getChar ()) != -1){
+
+ //
+ // No need to parse full string when parsing lambda arguments
+ //
+ if (lambda_arguments_parsing)
+ return Token.LITERAL_STRING;
+
+ while ((c = get_char ()) != -1){
if (c == '"'){
- if (quoted && peekChar () == '"'){
+ if (quoted && peek_char () == '"'){
string_builder.Append ((char) c);
- getChar ();
+ get_char ();
continue;
} else {
val = string_builder.ToString ();
}
if (!quoted){
- c = escape (c);
+ int surrogate;
+ c = escape (c, out surrogate);
if (c == -1)
return Token.ERROR;
+ if (surrogate != 0) {
+ string_builder.Append ((char) c);
+ c = surrogate;
+ }
}
string_builder.Append ((char) c);
}
}
if (res == Token.PARTIAL) {
+ if (parsing_block > 0) {
+ val = new LocatedToken (Location, "partial");
+ return Token.IDENTIFIER;
+ }
+
// Save current position and parse next token.
PushPosition ();
int next_token = token ();
bool ok = (next_token == Token.CLASS) ||
(next_token == Token.STRUCT) ||
- (next_token == Token.INTERFACE);
+ (next_token == Token.INTERFACE) ||
+ (next_token == Token.VOID);
PopPosition ();
- if (ok)
+ if (ok) {
+ if (next_token == Token.VOID) {
+ if (RootContext.Version == LanguageVersion.ISO_1 ||
+ RootContext.Version == LanguageVersion.ISO_2)
+ Report.FeatureIsNotAvailable (Location, "partial methods");
+ } else if (RootContext.Version == LanguageVersion.ISO_1)
+ Report.FeatureIsNotAvailable (Location, "partial types");
+
return res;
+ }
- if (next_token < Token.LAST_KEYWORD)
- Report.Error (267, Location, "The `partial' modifier can be used only immediately before keyword `class', `struct', or `interface'");
+ if (next_token < Token.LAST_KEYWORD) {
+ Report.Error (267, Location,
+ "The `partial' modifier can be used only immediately before `class', `struct', `interface', or `void' keyword");
+ return token ();
+ }
val = new LocatedToken (Location, "partial");
return Token.IDENTIFIER;
return res;
}
- private int consume_identifier (int s, bool quoted)
+ private int consume_identifier (int c, bool quoted)
{
- int pos = 1;
- int c = -1;
-
- id_builder [0] = (char) s;
+ int pos = 0;
- current_location = new Location (ref_line, Col);
+ if (c == '\\') {
+ int surrogate;
+ c = escape (c, out surrogate);
+ if (surrogate != 0) {
+ id_builder [pos++] = (char) c;
+ c = surrogate;
+ }
+ }
+
+ id_builder [pos++] = (char) c;
+ current_location = new Location (ref_line, hidden ? -1 : Col);
- while ((c = getChar ()) != -1) {
+ while ((c = get_char ()) != -1) {
loop:
if (is_identifier_part_character ((char) c)){
if (pos == max_id_size){
id_builder [pos++] = (char) c;
// putback_char = -1;
} else if (c == '\\') {
- c = escape (c);
+ int surrogate;
+ c = escape (c, out surrogate);
+ if (surrogate != 0) {
+ if (is_identifier_part_character ((char) c))
+ id_builder [pos++] = (char) c;
+ c = surrogate;
+ }
goto loop;
} else {
// putback_char = c;
//
// Optimization: avoids doing the keyword lookup
- // on uppercase letters and _
+ // on uppercase letters
//
- if (!quoted && (s >= 'a' || s == '_')){
+ if (id_builder [0] >= '_' && !quoted) {
int keyword = GetKeyword (id_builder, pos);
if (keyword != -1) {
val = Location;
if (val != null) {
val = new LocatedToken (Location, (string) val);
if (quoted)
- escapedIdentifiers.Add (val);
+ escaped_identifiers.Add (val);
return Token.IDENTIFIER;
}
}
val = new LocatedToken (Location, (string) val);
if (quoted)
- escapedIdentifiers.Add (val);
+ escaped_identifiers.Add (val);
return Token.IDENTIFIER;
}
// Whether we have seen comments on the current line
bool comments_seen = false;
-
val = null;
- for (;(c = getChar ()) != -1;) {
- if (c == ' ' || c == '\t' || c == '\f' || c == '\v' || c == 0xa0)
+ for (;(c = get_char ()) != -1;) {
+ if (c == '\t'){
+ col = ((col + 8) / 8) * 8;
+ continue;
+ }
+
+ if (c == ' ' || c == '\f' || c == '\v' || c == 0xa0 || c == 0)
continue;
if (c == '\r') {
- if (peekChar () == '\n')
- getChar ();
+ if (peek_char () != '\n')
+ advance_line ();
+ else
+ get_char ();
any_token_seen |= tokens_seen;
tokens_seen = false;
// Handle double-slash comments.
if (c == '/'){
- int d = peekChar ();
+ int d = peek_char ();
if (d == '/'){
- getChar ();
- if (RootContext.Documentation != null && peekChar () == '/') {
- getChar ();
+ get_char ();
+ if (RootContext.Documentation != null && peek_char () == '/') {
+ get_char ();
// Don't allow ////.
- if ((d = peekChar ()) != '/') {
+ if ((d = peek_char ()) != '/') {
update_comment_location ();
if (doc_state == XmlCommentState.Allowed)
handle_one_line_xml_comment ();
warn_incorrect_doc_comment ();
}
}
- while ((d = getChar ()) != -1 && (d != '\n') && d != '\r')
- if (d == '\n'){
- }
+ while ((d = get_char ()) != -1 && (d != '\n') && d != '\r');
+
any_token_seen |= tokens_seen;
tokens_seen = false;
comments_seen = false;
continue;
} else if (d == '*'){
- getChar ();
+ get_char ();
bool docAppend = false;
- if (RootContext.Documentation != null && peekChar () == '*') {
- getChar ();
+ if (RootContext.Documentation != null && peek_char () == '*') {
+ get_char ();
update_comment_location ();
// But when it is /**/, just do nothing.
- if (peekChar () == '/') {
- getChar ();
+ if (peek_char () == '/') {
+ get_char ();
continue;
}
if (doc_state == XmlCommentState.Allowed)
Location start_location = Location;
- while ((d = getChar ()) != -1){
- if (d == '*' && peekChar () == '/'){
- getChar ();
+ while ((d = get_char ()) != -1){
+ if (d == '*' && peek_char () == '/'){
+ get_char ();
comments_seen = true;
break;
}
}
is_punct_label:
- current_location = new Location (ref_line, Col);
+ current_location = new Location (ref_line, hidden ? -1 : Col);
if ((t = is_punct ((char)c, ref doread)) != Token.ERROR){
tokens_seen = true;
if (doread){
- getChar ();
+ get_char ();
}
return t;
}
if (c == '.'){
tokens_seen = true;
- int peek = peekChar ();
+ int peek = peek_char ();
if (peek >= '0' && peek <= '9')
return is_number (c);
return Token.DOT;
continue;
bool directive_expected = false;
- while ((c = getChar ()) != -1) {
+ while ((c = get_char ()) != -1) {
if (col == 1) {
directive_expected = true;
} else if (!directive_expected) {
return consume_string (false);
if (c == '\''){
- c = getChar ();
+ c = get_char ();
tokens_seen = true;
if (c == '\''){
error_details = "Empty character literal";
Report.Error (1010, Location, "Newline in constant");
return Token.ERROR;
}
- c = escape (c);
+
+ int surrogate;
+ c = escape (c, out surrogate);
if (c == -1)
return Token.ERROR;
- val = new System.Char ();
+ if (surrogate != 0)
+ throw new NotImplementedException ();
+
val = (char) c;
- c = getChar ();
+ c = get_char ();
if (c != '\''){
error_details = "Too many characters in character literal";
Report.Error (1012, Location, error_details);
// Try to recover, read until newline or next "'"
- while ((c = getChar ()) != -1){
+ while ((c = get_char ()) != -1){
if (c == '\n'){
break;
}
}
if (c == '@') {
- c = getChar ();
+ c = get_char ();
if (c == '"') {
tokens_seen = true;
return consume_string (true);
}
}
+ if (c == EvalStatementParserCharacter)
+ return Token.EVAL_STATEMENT_PARSER;
+ if (c == EvalCompilationUnitParserCharacter)
+ return Token.EVAL_COMPILATION_UNIT_PARSER;
+ if (c == EvalUsingDeclarationsParserCharacter)
+ return Token.EVAL_USING_DECLARATIONS_UNIT_PARSER;
+
error_details = ((char)c).ToString ();
return Token.ERROR;
private void handle_one_line_xml_comment ()
{
int c;
- while ((c = peekChar ()) == ' ')
- getChar (); // skip heading whitespaces.
- while ((c = peekChar ()) != -1 && c != '\n' && c != '\r') {
- xml_comment_buffer.Append ((char) getChar ());
+ while ((c = peek_char ()) == ' ')
+ get_char (); // skip heading whitespaces.
+ while ((c = peek_char ()) != -1 && c != '\n' && c != '\r') {
+ xml_comment_buffer.Append ((char) get_char ());
}
if (c == '\r' || c == '\n')
xml_comment_buffer.Append (Environment.NewLine);
if (current_comment_location.IsNull) {
// "-2" is for heading "//" or "/*"
current_comment_location =
- new Location (ref_line, col - 2);
+ new Location (ref_line, hidden ? -1 : col - 2);
}
}
public void cleanup ()
{
if (ifstack != null && ifstack.Count >= 1) {
- current_location = new Location (ref_line, Col);
+ current_location = new Location (ref_line, hidden ? -1 : Col);
int state = (int) ifstack.Pop ();
if ((state & REGION) != 0)
Report.Error (1038, Location, "#endregion directive expected");